A Cover-based Enhanced Direct Search Approach for Contact Detection of Convex Polyhedral Blocks and Its Application with 3D DDA

Abstract

In implicit numerical simulation of block systems, the establishment of complete inequality constraints and the accurate enforcement of contact constraints are fundamental issues. However, the conventional direct search (DS) method is usually based on a local type identification rule to establish contact pairs, which may lead to missed and falsely identified contact pairs when applied to irregular polyhedral blocks. In this study, a cover-based enhanced direct search (CDS) approach for arbitrary convex polyhedral blocks is proposed to achieve robust contact analysis of block system with irregular shaped blocks. Based on the concepts of the entrance block and contact covers, contact identification rules and global search loops which consider geometric tolerances per calculation step are established for vertex-to-face and cross edge-to-edge contacts. A classification and treatment strategy is then developed for valid contact pairs and potential contact pairs according to their interpenetration states. In addition, a multiple contact pair resolving algorithm and a modified open-close iteration (OCI) scheme are proposed to resolve contact plane indeterminacy and the incompleteness of entrance path determination criteria. The proposed approach is implemented in a 3D DDA framework and validated through six numerical examples. The results demonstrate that the CDS approach effectively eliminates the robustness deficiencies of the local type identification-based DS method, while exhibiting stronger robustness and higher computational efficiency in the contact analysis of complex convex polyhedral block systems.